Humans have to kind of go through all these kinds of policies. Now, machines do, they give humans a recommendation. And eventually, then machines are going to be talking with our customers on the front line as well. And then on the delivery marketplace and the mobility marketplace, we’ve been using advanced machine learning algos for routing, for matching, for pricing, all of these algos generally improve on a year-on-year basis and accrue hundreds of millions of dollars of either incremental bookings that call it VC neutral or an improved chain algorithms to reduce, let’s say, cost per transaction on the delivery space. So, these are just some of the areas where ML is working out. And listen, for us, ML is a powerful technology, period. But because we’re the largest player in the world, we are gathering more data for more customers across a wider range of behaviors than anybody else.
So, we think in a world where ML becomes more important Uber becomes competitively stronger, so to speak, because of the set of data that we have that’s really unmatched and unrivaled, including customers who are engaging both in the mobility and the delivery marketplace. So, it’s pretty powerful. In terms of the other use cases that we’re seeing, listen, Uber for Business is something that’s very, very promising. We continue to penetrate into the corporate space. It’s great to see companies now get their travelers on the road again and we definitely saw an uptake there. But with Uber for Business, we’re also looking at verticals like health and transit as well that are very, very healthy and hold a ton of promise. For me, travel is something that’s very much in my heart because of my old Expedia days.
And we’ve talked about travel being a very, very strong kind of signal for Uber and travelers typically using Uber because of our global footprint, kind of perspective here. In 2022, nearly 700 million trips globally were taken by consumer that’s out of their primary city, which is pretty amazing. And every quarter, about 20% of our users took a domestic trip outside of their primary cities. So our users, they’re high-income earners, they tend to travel a lot and that becomes a good segment for us to target. We’ve done so in the UK where we’re quite optimistic about our travel business in the UK and some of the early signals that we see, for example, in the UK, you can book trains and buses, 60% of train and coach users and 25% of our flight users are now repeat users on Uber as well.
So they try the product because we have a big audience. But more importantly, they’re coming back to the product, which is great. So we got multiple avenues for growth and travel and business are just one amongst many honestly.
Brian Nowak: Okay. Thanks, Dara.
Dara Khosrowshahi: Yeah. Next question.
Operator: Our next question comes from Eric Sheridan with Goldman Sachs. Please go ahead.
Eric Sheridan: Thanks so much. And first, thanks to Nelson for all the insights and conversations over the years. Thanks. I wish you best going forward. Maybe Dara, I think two bigger picture ones for you. One on Uber One. As it continues to scale globally and the base of subscribers continues to build, what are some of the early learnings that have now translated into scale about how you’re thinking about bringing that subscription to market more globally than in the early years and what that might mean for the business in the years ahead? That’s number one. And then on driver supply. We were in a very different position a year ago where you were using a lot of incentives on driver supply. As we come out of that period and things have normalized, how are you thinking about optimizing driver supply for efficiency cost, maybe even insurance? So we should be thinking about that as an initiative going forward? Thanks so much.
Dara Khosrowshahi: Sure, Eric. So as far as Uber One goes and lessons learned there, honestly, the Uber One customer behavior has been pretty darn consistent in that Uber One consumers spend four times the amount that nonmembers do monthly basis. And retention is more than 15% higher for members versus nonmembers. That pattern has remained. There were some question as to, A, as we expand the user base, is that ForEx going to come down to a 3x, for example, but that hasn’t happened. The 4x and the 15% retention have continued. The focus now for us with Uber One, I’d say is threefold. One is keep expanding geographically. And we just introduced Uber One into a couple of more markets as well. So now we’ve got 15 million members across 18 countries.
Number two is really focused on Uber One retention. And when you look at the Uber One benefits, the benefits are typically monetary in nature. So you get discounts on your food, you get delivery for free. You get cash back on mobility. What we want to do going forward is also provide benefits that are non-monetary in nature, let’s say, advanced matching upgrades to different cars or head of the queue matching when you get into an airport, where there aren’t that many cars around. Those are more nonmonetary benefits that we think that our members will very much appreciate. And the design spec there is to drive member retention. Retention levels are high, but they can always get higher, and we always want to make sure that we’re not kind of driving just member growth, we also want to drive retention as well.
And the third area is really optimizing around our mobility use case. Delivery already members account for more than 40% of bookings with mobility, it’s in the mid-teens. It’s much earlier in terms of penetration. And I think we can do a lot more in terms of the member experience for mobility users and just continuing to optimize that. Turning to the driver supply question. Our supply position is the best that it’s been. We’ve got over 6.5 million earners. And I would say those earners are actually earning this quarter, $15.9 billion, which is up 23% on a constant currency basis. So we’re really proud of our being one of the largest earnings, if not the largest earnings platform in the world in a flexible way and the fact that earnings are actually going up faster than gross bookings, which is something that we’re proud of.
